University of the Witwatersrand Institute for Advanced Social Research

UNIVERSITY OF THE WITWATERSRAND INSTITUTE FOR ADVANCED SOCIAL RESEARCH

Seminar Paper to be presented in RW 7003 SEMINAR ROOM AT 4.00pm 29th AUGUST 1994

Title: "If we can't call it the mfecane, then what can we call it?": Moving the debate forward.

by: John Wright

No 364 The mfecane as fetish In the last six years a major controversy has blown up among historians of southern Africa about the historical reality or otherwise of the phenomenon commonly known as the mfecane.1 Since it was first popularized by John Omer-Cooper in his book The Zulu Aftermath, published in 1966,2 the term has become widely used as a designation for the wars and migrations which took place among African communities across much of the eastern half of southern Africa in the 1820s and 1830s. For more than a century before Omer-Cooper wrote, these upheavals had been labelled by writers as 'the wars of Shaka' or 'the Zulu wars'; today the view remains deeply entrenched among historians and public alike that the conflicts of the period were touched off by the explosive expansion of the Zulu kingdom under Shaka. In a chain reaction of violence, so the story of the mfecane goes, warring groups carried death and destruction from the Zululand region southwards into Natal and the eastern Cape, westward onto the highveld, and northwards to the Limpopo river and beyond. The violence came to an end only when most of the communities which had managed to survive the supposed chaos of the times had been amalgamated into a number of large defensive states under powerful kings.

Since Julian Cobbing published his now well-known critique of the concept of the mfecane in 1988,3 a number of other historians, myself among them, have joined him in challenging the idea that the concept has any descriptive or analytical value. These historians do not deny that the 1820s and 1830s were a time of often fierce conflict in much of southern Africa, but they regard the notion of the mfecane as a fundamental obstacle to understanding the history of the sub-continent in the late eighteenth and early nineteenth centuries.4

1. This paper draws on an earlier one, entitled 'Raiding, trading and political change: the central and eastern regions from the 1760s to the 1830s', which I presented to the Second Colloquium on School History Textbooks for a Democratic South Africa held in Rustenburg in November 1993. 2. J.D. Omer-Cooper, The Zulu Aftermath: A Nineteenth Century Revolution in Bantu Africa, London, 1966. 3. J. Cobbing, 'The mfecane as alibi: thoughts on Dithakong and Mbolompo', Journal of African History, vol. 29 (1988), pp. 487-519. 4. Many of the debates which have emerged since 1988 round the concept of the mfecane are taken up in the collection of essays currently being edited by Carolyn Hamilton under the title, The Mfecane Aftermath. Publication jointly by Witwatersrand University Press and Natal University Press is due early in 1995. My thanks to the editor and pub- lishers for permission to see pre-publication proofs of this work. As I see it, there are six particular sets of objections that can be raised against the notion of the mfecane. First, it serves to maintain the long-established segregation from each other of histories which, it is increasingly clear to some historians, were becoming more and more intertwined from at least the mid-eighteenth century onwards: that is, the history of the white-dominated societies of the expanding Cape colony, and the history of the black societies of the southern African interior. Second, it lumps together histories which need to be treated separately: for example, the history of the Natal region in the later eighteenth century, of the southern highveld in the 1820s, of central Mozambique in the 1830s, of Malawi in the 1840s. Third, it is voraciously and irredeemably teleological. It has long since swallowed up the history of the Zululand region in the later eighteenth century by portraying it as little more than a preface to the postulated 'Zulu explosion' of the 1820s. In Omer-Cooper's latest treatment of it, it is similarly threatening also to absorb the pre-1820s history of the western Transvaal, the eastern Transvaal, Natal-Mpondoland, and the Cape northern frontier region.5 Fourth, it makes for a heavily distorted periodization of political change by focussing mainly on the 1820s and 1830s, and largely neglecting previous decades. Fifth, it unproblematically attributes the often violent political transformations of the 1820s and 1830s primarily to the expansion of the Zulu kingdom, and underplays or neglects wider sets of causes. In particular, it has virtually nothing to say about the impact of forces rooted in the expansion of the frontiers of European trading, raiding and settlement. Sixth, it greatly exaggerates the level of violence and destruction that took place in the 1820s and 1830s. In one account after another, words like 'chaos', 'turmoil', 'bloodshed', 'massacres', 'holocaust' are used without reflec- tion to describe the events of the period.

Critics of the concept of the mfecane are by no means united in their attempts to provide alternative explanations of the political changes which were taking place in the interior and eastern regions of southern Africa from the mid-eighteenth century onward. They also argue from widely divergent per- spectives about the meanings that can be read into particular sources of evidence. Often fierce criticisms of aspects of their arguments, particularly of Cobbing's, have been made by other historians.5 This is not the place to spell out in the detail the various lines of debate which have emerged. But we need to register that one of the effects of these debates,

5. J.D. Omer-Cooper, 'Has the Mfecane a future? A response to the Cobbing critique'. Journal of Southern African Studies r vol. 19, no. 2 (1993), pp. 273-94; J.D. Omer-Cooper, His- tory of Southern Africa. 2nd ed., Cape Town, 1994, pp. 52- 68. 6. See Hamilton, ed., Mfecane Aftermath, in press. both written and spoken, has been to leave numbers of readers and listeners in a state of deep confusion. This is hardly surprising when the participants themselves are not agreed what they are arguing about. Critics of the mfecane want to abandon the notion altogether, and move away from what they see as its inescapably Zulu-centric genre of explanations (even if, at this stage of the debate, they are not always able to expunge the word from the titles of their papers). Some critics of the critics, like Jeff Peires, con- tinue to feel that the main issue has to do with the rise of the Zulu state, and want to hold on to the established notion of the mfecane.7 Others, like Omer-Cooper, seek to define a new kind of mfecane so as to accommodate some of the chal- lenges that have been made to the old one.8 Other authors, more sympathetic to the critics, recognize the need to move on from Zulu-centric interpretations but still want to retain the term mfecane or difaqane as a generic designation for the upheavals of the 1820s and 1830s.9

Confusions are compounded by the inclination of some of Cobbing's critics to dismiss holus bolus what they call 'the Cobbing hypothesis' because of the major flaws in his use of evidence. Here they fail to distinguish between a general and a particular hypothesis which emerge from his arguments. The former, which in my view constitutes a major contribution to South African historiography, and is difficult to gainsay, has to do with his identification of European commercial and colonial expansion as having been a major factor in the history of the interior and east from at least the mid-eighteenth century onward. The latter, in criticisms of which I concur, has to do with the heavy emphasis which he places on the role of European slave-raiding. It is perfectly possible, in fact necessary in my view, to query the latter while holding on to the former. Against this background of disagreement about the nature of the disagreements on the mfecane,10 it is understandable that after one particular seminar a historian colleague could turn to me with the question which I have incorporated into the title of this paper, 'If we can't call it the mfecane, then

7. J.B. Peires, 'Paradigm deleted: the materialist interpretation of the Mfecane', Journal of Southern African Studies, vol. 19, no. 2 (June 1993), pp. 295-313. 8. See note 4 above; also the critical comments on Omer- Cooper's revised views in my essay, 'Beyond the concept of the "Zulu explosion": comments on the current debate', in Hamilton, ed., The Mfecane Aftermathf in press. 9. Neil Parsons, A New History of Southern Africa. 2nd ed. , London, 1993, p. vi & chs. 5 & 6. 10. For more comments on this point, see Carolyn Hamilton's introductory essay to The Mfecane Aftermath. what can we call it?' My answer to this question takes two forms. The first is simply that there is no single 'it' out there to give a name to in the first place. The issue is not about finding a new name for a particular historical 'event' or set of events. Nor is it about finding new causes for such an event. It has to do with abandoning not just the name but the whole concept of the mfecane. In this connection, it would, I suggest, be highly instructive to consider the effects of what Edwin Wilmsen in another context has called 'the need to name'.11 We need to ask why, since at least the mid-nineteenth century, so many students of southern African history have had a powerful and deep-seated need to see an 'it' in the upheavals of the 1820s and the 1830s, and to give that 'it' a name - the wars of Shaka, the Zulu wars, and now, more singularly than ever, the mfecane. Why is it that they want to portray a multiplicity of disparate happenings, spread over many years and taking place in widely separated places, as a unitary, bounded 'event', a discrete 'thing' in the past.? Even historians who are prepared to concede the validity of many of the criticisms that have been mounted against the concept of the mfecane, and to abandon this specific term, find difficulty in letting go of the idea that the upheavals of the 1820s and 1830s constitute a single phenomenon for which there is a single line of explanation. This is not the place to try to answer the question of why this is so: the history of the concept of the mfecane needs its own separate study. My concern, rather, is to make the point that to understand the issues at stake in current debates on the mfecane we need to be aware of how, over a long period of time, certain events of the 1820s and 1830s have become objectified, fetishized even, into a nameable 'fact' of southern African history.

The second form that my answer to the question posed in the title takes is constituted by the rest of this paper. While the debates round Cobbing's hypotheses, and about the nature of the evidence on which they are based, have pointed to the urgent need for a complete reassessment of the history of the interior and eastern regions of southern Africa in the late eighteenth and early nineteenth centuries, they have not yet taken us much towards new conceptualizations of this history, and new lines of research. It is time for historians who reject the concept of the mfecane to move on from the Cobbing- centric arguments of the last six years, and get on with the business of articulating and researching empirical alterna- tives to mfecane-informed history. What follows is an outline of a possible alternative approach to explaining the political transformations which took place in the territories north of the Orange river and east of the Kalahari desert in the period from the 1760s to the 1830s. In a reversal of the usual

11. E.N. Wilmsen, Land Filled with Flies: a Political Economy of the Kalaharif Chicago & London, 1989, p. 26. pattern of mfecane-informed history, I deal first with the history of the interior, and then with that of the eastern regions. My aim in doing this is to enable the significance of forces for change emanating from the expansion of the Cape colony to be appreciated, and, concomitantly, for the emergence of the Zulu kingdom to be seen in a wider context than usual. The interior: the expansion of trade after 1760 Historians have long recognized that the later eighteenth century saw significant political changes among the societies of the region between the northern and central Drakensberg and the Indian Ocean. They have often described these changes as leading towards 'state formation', by which is meant the processes in which numbers of relatively small-scale, uncentralized and unstratified chiefdoms became amalagamated into relatively large-scale, centralized and stratified polities, or states. It is only recently that historians have begun to recognize - perhaps not yet fully - that similar, if less pronounced, processes were taking place at much the same time among the Tswana-speaking societies on the south-east fringes of the Kalahari desert. To a large extent this lag has been a product, of the relative paucity of source material on the history of the interior before the 1820s, and of the difficulties of interpreting what evidence there is.

Our knowledge of the history of the interior regions before the second half of the eighteenth century depends very largely on oral traditions which were written down in the nineteenth and twentieth centuries by missionaries, officials and academ- ics, nearly all of them white. Most of these traditions tell stories of specific migrations, wars and disputes over chief- ship, and it is difficult to work out from them the broad patterns of the region's politics. Historians today still draw heavily on the syntheses of tradition produced from the 1930s onward by anthropologist Isaac Schapera, but, like many scholars then and since, Schapera tended to take traditions too much at face value as sources of evidence, and his histories stand in need of critical re-analysis. The same point could be made of the more recent exegeses produced by Martin Legassick and by Neil Parsons, both of which remain essential reading.12 However, the consensus among modern-day students of Tswana history that the final decades of the eighteenth century were a period of increased conflict and political expansion among the southern Tswana chiefdoms seems well- enough founded. The most important of these polities appear

12. M. Legassick, 'The Sotho-Tswana peoples before 1800', in L. Thompson, ed., African Societies in Southern Africaf London, 1969, ch. 5; Parsons, New History of Southern Afri- ca. 2nd ed., ch. 3. to have been those of the Ngwaketse, the Hurutshe, the Kwena, the Kgatla, the Rolong and the Tlhaping.13 It seems to me highly significant that these chiefdoms were situated at or near the intersection of several well- established long-distance trade routes. Evidence from Euro- pean travellers who were reaching this region by the 1810s and 1820s indicates that one of these led eastwards to the Indian Ocean 800 kilometres away, and another south-westwards to the Orange river valley and the Cape frontier regions.14 Accord- ing to the findings of recent archaeological research, goods from Muslim and Portuguese traders on the east coast had been reaching the interior, if in small quantities, for many centuries before this.15 Snippets of information recorded by European officials and travellers from Khoekhoe informants in the seventeenth and eighteenth centuries suggest that trade between the southern Tswana and Khoekhoe communities to the south had also been in existence for a long time.16

A third long-distance route may have led from the southern Tswana region to the north and north-west. Archaeological evidence cited by Wilmsen indicates that goods acquired from Portuguese traders in Angola were reaching what is now north- western Botswana by the early seventeenth century,17 and it is possible that in due course this trade may have extended to the eastern and south-eastern edges of the Kalahari.

13. See the works cited in the previous note; also Andrew Manson, 'The Hurutshe in the Marico district of the Trans- vaal, 1848-1914', unpubl. Ph.D. thesis, University of Cape Town, 1990, pp. 42-53. 14. M. Wilson, 'The Sotho, Venda and Tsonga', in M. Wilson & L. Thompson, eds., The Oxford History of South Africa, vol. 1, Oxford, 1969, pp. 149-53. 15. The evidence is summarized in T. Maggs & G. Whitelaw, ' A review of recent archaeological research on food-producing communities in southern Africa', Journal of African His- tory, vol. 32 (1991), pp. 14-15. 16. C. Saunders, 'Early knowledge of the Sotho: seventeenth and eighteenth century accounts of the Tswana', Quarterly Bull. S.A. Library, vol. 20 (1965-6), pp. 63-4; R. Elphick, Kraal and Castle: Khoikhoi and the Founding of White South Africa. New Haven & London, 1977, pp. 65, 67; G. Okihiro, 'Precolonial economic change among the Tlhaping, c.1795- 1817', Internat. Jnl. Afr. Hist. Studies, vol. 17 (1984), pp. 70-71; G. Harinck, 'Interaction between Xhosa and Khoi: emphasis on the period 1620-1750', in Thompson, ed. , Afri- can Societies, pp. 160, 164.

17. Wilmsen, Land Filled with Fliesr pp. 76, 88; E.N. Wilmsen, 'For "trinkets" such as beads: Khoisan participation in the European Atlantic trade', unpubl. conference paper, Tutz- ing, Germany, July 1994 (my thanks to the author for permission to cite this paper). Evidence from early travel accounts and from archaeological research cited by recent commentators indicates that commodities imported into the interior from maritime sources were largely prestige items like beads, cloth and cowrie shells. In exchange, traders on the coast required ivory, copper, skins and pelts. There is no evidence that goods were carried directly between the coast and the interior: rather, they passed through numerous local trading networks in which both foreign and locally produced goods were exchanged.18 The volume of goods traded was never high, but it is likely that their value was enough to encourage chiefs to try to keep control over trade as far as possible for themselves. Disputes over trade were probably a frequent cause of conflict between chiefs and local community leaders, and between different chiefdoms.

Martin Legassick's researches of more than twenty-five years ago on the history of the Cape northern frontier region indicate that in the later eighteenth century the volume of trade along the route between the southern Tswana chiefdoms and the Khoekhoe communities to the south was beginning to expand. By the 1760s, the frontier of Cape colonial settlement had reached the Nuweveld and Sneeuberg mountains, and Dutch stock- farmers were seasonally grazing their livestock further north towards the Orange river. More and more they came into contact with the groups of Khoekhoe pastoralists 'who lived along the river. Though it was illegal for colonists to trade across the colonial boundary, frontier farmers traded beads, firearms and horses to the Khoekhoe in exchange for cattle. In turn, the Khoekhoe exchanged beads and other manufactures with the Tlhaping to the north of the Orange in return for cattle. With the Khoekhoe and the Tlhaping as middlemen, goods from the Cape colony began to move into southern Tswana chiefdoms in increased quantity.^

At much the same time, the we 11-documented researches of Alan Smith and David Hedges tell us, there was a marked expansion of the ivory trade at Delagoa Bay in southern Mozambique.20

18. For general comments on trade among the southern Tswana to the early nineteenth century, see M.C. Legassick, 'The Gri- qua, the Sotho-Tswana, and the missionaries, 1780-1840: the politics of a frontier zone', unpubl. Ph.D. thesis, UCLA, 1969, pp. 225-41; Okihiro, 'Precolonial economic change among the Tlhaping', IJAHS. vol. 17 (1984), pp. 68-78; A. Manson, 'The Hurutshe and the formation of the Transvaal State, 1835-1875', IJAHS, vol. 25 (1992), pp. 87-8. 19. Legassick, 'The Griqua', unpubl. thesis, pp. 126-33. 20. A. Smith, 'The trade of Delagoa Bay as a factor in Nguni politics, 1750-1835', in Thompson, ed., African Societiesr ch. 8; A. Smith, 'The struggle for control of southern Mocambique, 1720-1835', unpubl. Ph.D. thesis, UCLA, 1970, pp. 162-5, 190-5; D. Hedges, 'Trade and politics in southern Mozambique and Zululand in the eighteenth and ear- From the 1760s onwards, English and Indian merchants were coming from India in increasing numbers to buy ivory from the chiefs round the bay in exchange for cloth, beads, and metal. Though it was chiefdoms nearer the bay which were most directly affected by the growth of this trade, some historians have suggested that its impact was felt in the southern Tswana region as well.21 Also in this period, trade between southern Tswana chiefdoms and others further north may have been expanding. Wilmsen brings to our attention that Portuguese trading activity in southern Angola was expanding at this time, and that by the 1770s beads from this commerce were reaching as far south as the Orange river.22 Among the Tswana on the eastern edges of the Kalahari, Parsons argues, ivory and furs from north-west Botswana were in increasing demand in these years.23

The interior: the expansion of the 'firearms frontier' It was not only through the expansion of trade from the south- west and from the east that influences from the outside world were felt among the southern Tswana chiefdoms in the later eighteenth century. As Legassick's seminal and richly tex- tured works have long since revealed, the advance of the Cape colonial frontier of settlement brought closer the frontier of raiding.24 Official and unofficial commandos marauded northwards for livestock and for Bushman women and children, whom the Dutch colonists wanted as labourers on their stock farms. (The term 'Bushmen' is here used as a designation for the members, most of them probably Khoisan-speaking, of fluid, unstable, impermanent groupings which lived by varying combinations of hunting, gathering, herding, trading and raiding.) As the area dominated by the Dutch expanded, groups of Khoekhoe and Bushmen, many of which were already in conflict with the frontier farmers, were sgueezed out into the arid regions along the middle and lower Orange river. Here they came into competition and conflict with other Khoekhoe and Bushman groups which were already living there.

ly nineteenth centuries', unpubl. Ph.D. thesis, London Uni- versity, 1978, chs. 3, 6. 21. Parsons, New History, p. 48; P. Delius, The Land Belongs to Us: the Pedi Polity, the Boers and the British in the Nineteenth Century Transvaal. Johannesburg, 1983, p. 18.

22. Wilmsen, Land Filled with Fliesr pp. 88, 91, 95. 23. N. Parsons, 'Pre-difaqane wars in the west', unpubl. conference paper, University of the Witwatersrand, 1991, p. 3. 24. Legassick, 'The Griqua', unpubl. thesis, chs. 2, 3; M. Legassick, 'The northern frontier to c.1840: the rise and decline of the Griqua people', in R. Elphick & H. Giliomee, eds., The Shaping of South African Society. 1652-1840r Cape Town, 1989 (1st ed. 1979), ch. 8. Recent studies by Legassick and by Nigel Penn, using official archival documents, unpublished missionary records and published travel accounts as sources, complement older work by Legassick himself and by Shula Marks in charting the impact of the expansion of the frontier on the indigenous inhabitants.25 The prime effect was the breakdown of established pastoralist and hunting and gathering economies, and the emergence of mixed raiding bands of Khoekhoe, Bushmen, and Tswana, many of them equipped with guns and horses. Some of these groups were joined by runaway slaves and by bandits and deserters, both white and 'coloured', from the Cape colony. They raided livestock from one another, and from the Dutch to the south. By the 1780s, at least, they were also making depredations on Tswana communities to the north.26 The researches of Miklos Szalay and, more recently, of Elizabeth Eldredge, go far beyond the older frontier studies of P.J. van der Merwe in revealing the extent to which these bands also raided for women and children from weaker Bushman communities, and sold them as slaves to the Dutch.27 (In reaction to Cobbing's slaving hypothesis, some historians have insisted that not all labourers on Boer farms were slaves, and that the status of Bushmen in Boer-dominated society ranged from that of slaves through various degrees of serfdom to that of clients. The point is an important one, but it seems to me difficult to call those labourers whoe were seized and kept in servitude by force anything other than slaves.) By the early years of the nineteenth century, these groups, which were collectively known as Kora or Korana, were a major destabilizing force along the middle and lower Orange. Though direct evidence on

25. Legassick, 'The Griqua', unpubl. thesis, pp. 60-71, 133- 41, 241-7; Legassick, 'The northern frontier', in Elphick & Giliomee, eds., Shaping, pp. 368-76; S. Marks, 'Khoisan resistance to the Dutch in the seventeenth and eighteenth centuries', Jnl. Afr. Hist.. vol. 13 (1972), pp. 72-9; N. Penn, 'Labour, land and livestock in the Western Cape during the eighteenth century: the Khoisan and the colonists', in W. James & M. Simons, eds.. The Angry Divide: Social and Economic History of the Western Caper Cape Town, 1989, pp. 13-18. 26. Legassick, 'The northern frontier', in Elphick & Giliomee, eds., pp. 369, 374. 27. M. Szalay, Ethnologie und Geschichte: Zur Grundlegung einer Ethnologischen Geschichtsschreibung. Berlin, 1983, ch. 6 (my thanks to the author for making the English translation of this chapter, which he plans to publish, available to me); E. Eldredge, 'Slave-raiding across the Cape frontier', in E. Eldredge & F. Morton, eds., Slavery in South Africa: Captive Labor on the Dutch Frontier, Boulder, Westview Press, in press, ch. 4 (my thanks to Elizabeth Eldredge for making a draft of this chapter available to me). 10 the subject is very thin, it is likely that they posed a con- tinual threat to Tswana stock-keeping communities to the north and north-east. Sometime towards the end of the 1790s a new element was added to the politics of the middle Orange region with the arrival of several groups of Griqua from lower down the river. These people were pastoralists, largely of Khoekhoe descent, who were seeking to escape from the disturbances that were being caused by the expansion of Dutch stock farmers towards the lower Orange. They settled along a line of springs near the junction of the Orange and Vaal in territory occupied by communities of Kora and Bushmen, numbers of whom had little option but to join them as clients.28

Like the Kora, the Griqua were equipped with guns and horses, and some groups took to raiding. The more important Griqua leaders, however, were interested less in raiding than in livestock-keeping, in hunting, and in trying to break into the well established trade between the Tswana and the Cape colony. Some groups even tried to cultivate crops. In 1801, one leader, seeking protection from rival groups, and also wanting to establish closer commercial links with the colonial world, arranged for missionaries from the Cape to begin work among his followers. By the 1820s, other missionaries had begun operations among certain of the southern Tswana chiefdoms. To support themselves, some of them became part-time traders. Several became influential political advisers to chiefs. From this period dates the beginning of the more or less con- tinuously documented history of the region north of the middle Orange.2^

28. Legassick, 'The Griqua', unpubl. thesis, pp. 146-8; J.S. Marais, The Cape Coloured People 1652-1937. Johannesburg, 1968 (1st publ. London, 1939), pp. 32-5; R. Ross, Adam Kok's Griauas, Cambridge, 1976, pp. 12-15. 29. On the Griqua and neighbouring societies in the first two decades of the nineteenth century, see Legassick, 'The Griqua', thesis, pp. 151-9, 163-8, & chs. 4 & 6; Legassick, 'The northern frontier', in Elphick & Giliomee, eds.. Shaping, pp. 376-90; R. Wagner, 'Coenraad de Buys in Transorangia', in The Societies of Southern Africa in the 19th and 20th Centuries, vol. 4, Institute of Commonwealth Studies, university of London, 1974, pp. 1-8; J. du Bruyn, 'Die aanvangsjare van die Christelike sending orider die Tlhaping', unpubl. M.A. thesis, Unisa, 1980, chs. 3 & 4; R. Beck, 'The legalization and development of trade on the Cape frontier, 1817-1830', unpubl. Ph.D. thesis, Indiana University, 1987, ch. 6; M. Kinsman, 'Populists and patriarchs: the transformation of the captaincy at Griqua Town, 1804-1822', in A. Mabin, ed., Organization and Eco- nomic Change. Southern African Studies vol. 5, Johannes- burg, 1989, pp. 1-20. 11 The history of the relatively settled Griqua communities has attracted its fair share of researchers. Much less well known is the contemporary history of small groups of Xhosa which, as a result of the increasing instability of the eastern frontier of the Cape, were quitting the region and establishing themselves along the middle Orange. Work by Peter Kallaway and Elizabeth Anderson has served to illuminate the role which they played in further destabilizing this already volatile region. They attracted followings of Khoekhoe, Bushmen, and refugees from the colony, and, with the firearms which they had acquired, raided widely in the region round the middle Orange and lower Vaal. Though they mostly avoided attacking the stronger of the Griqua groups, the raids which they made against the Tlhaping threatened to disrupt the trade which these groups were trying to establish to the northward.30

It is against this background of expanding trade to the south, east and north, and of increasing political instability in the regions to the south, that we need to place the history of the southern Tswana chiefdoms in the last two decades or so of the eighteenth century and the first two of the nineteenth. The main factors which led to conflict and expansion among them were, I would argue, a combination of rivalry for control of trade routes and ivory-producing areas, and the need for protection against Kora, Griqua, Xhosa and Dutch raiders from the south. In the south, the Tlhaping, in alliance with some Kora groups, competed for dominance with one of the Rolong chiefdoms. Further north, the Ngwaketse broke away from a Kwena group and established a separate chiefdom which expanded rapidly in size and strength. It came into conflict with both the Tlhaping and the Rolong, and with the Hurutshe chiefdom, which was also expanding.3*

It was probably during this period that there began the con- centration of settlement which produced the large southern Tswana towns remarked on by European visitors in the early decades of the nineteenth century. There has been a good deal of debate in the literature about their origins: recently archaeologist Tom Huffman has argued that the process most probably began in the eighteenth century, 'perhaps not before AD 1750', and that it was due primarily to defensive needs in a time of 'tremendous insecurity'. As I do in this paper,

30. P. Kallaway, 'Danster and the Xhosa of the Gariep: towards a political economy of the Cape frontier 1790- 1820', African Studies, vol. 41 (1982), pp. 143-60; E. Anderson, A History of the Xhosa of the Northern Cape 1795- 1879. Centre for African Studies, University of Cape Town, 1987, ch. 2; Legassick, 'The Griqua', thesis, pp. 248-50. 31. Legassick, 'The Griqua', thesis, pp. 251-80; Parsons, New History, pp. 48-50; Manson, 'The Hurutshe', unpubl. thesis, pp. 47-53; A. Manson, 'Conflict on the western highveld/southern Kalahari, c.1750-18 20', unpubl. conference paper, University of the Witwatersrand, 1991. 12 Huffman draws on Legassick's work to argue that this insecurity was caused primarily by raids from the south on the part of Griqua, Kora and Xhosa.32 By the early nineteenth century, when the first European travellers visited them, some of these towns had populations of ten or fifteen thousand. At these places, conflicts over grazing and agricultural land may have intensified. Periodic droughts were probably another factor in contributing to rising levels of social and political tension. These were not simply internal 'African' conflicts, for from the 1810s, at least, white gunmen like the well-known Coenraad de Buys, as well as white missionaries, were also playing a role in local politics.33 Parsons's work indicates that fighting was becoming fiercer as chiefs sought not only to seize cattle from rival polities, but also to conquer them and eliminate their leaders. He concludes, 'By 1820 warfare was so widespread among the western Tswana that hardly a chiefdom had not seen its chief recently killed in battle'.34 The implications of this statement, first made twelve years ago, have entirely escaped those historians who want to hold on to the notion of a Zulu-caused difaqane that supposedly began on the highveld only after 1820. There is a certain amount of evidence that intensified conflict among the southern Tswana chiefdoms spilt over south and and east into the territories across the Vaal river. This, I suggest, is one possible reading of evidence cited by D.F. El- lenberger in his History of the Basuto that in the early years of the nineteenth century the Taung were beginning to attack and subjugate neighbouring chiefdoms in what is now the northern and central Orange Free State.35 Further to the east, one of the Tlokwa chiefdoms also seems to have been growing in power, possibly, as Sanders suggests, because it was situated in a locality from which its leaders could control an east- west trade over the Drakensberg escarpment.36 Pressures on

32. T. Huffman, 'Archaeological evidence and conventional explanations of Southern Bantu settlement patterns', Africaf vol. 56 (1986), esp. pp. 291-3. The quotations are from p. 292. 33. On the role of white gunmen see Wagner, 'Coenraad de Buys', pp. 2-4; Manson, 'Conflict on the western highveld', unpubl. paper, p. 6. 34. Parsons, New History. 2nd ed., p. 52. 35. D.F. Ellenberger, History of the Basuto. Ancient and Mod- ern, London, 1912, p. 60. See also T. Maggs, Iron Age Com* munities of the Southern Highveld. Pietermaritzburg, 1976, pp. 309-10; Parsons, New History, 2nd ed., p. 52. 36. P. Sanders, Moshoeshoe: Chief of the Sotho, London, 1975, p. 19. On trade among the Sotho of the southern highveld in the early nineteenth century, see E. Eldredge, A South African Kingdom: the Pursuit of Security in Nineteenth- Century Lesotho. Cambridge, 1993, pp. 19-25. 13 the Sotho-speaking chiefdoms of the southern highveld came not only from the zone of conflict among the Tswana to the north- west, but also from what Legassick has called the 'firearms frontier' along the middle Orange to the south-west.37 By the 1810s Boer farmers were expanding towards the river in what is now the Colesberg area,38 and raiding by Boers, Kora and Gri- qua for cattle and slaves may by this time have been reaching as far as the Caledon valley. More and more, I suggest, the existing political order in the region was beginning to come under threat. The interior: the upheavals of the 1820s and 1830s In the early and mid-1820s, conflicts among the Sotho and Tswana chiefdoms of the highveld became more intense. The commonly held view is that this was because of the intrusion at this time of three groups from east of the Drakensberg. These were the Hlubi and the Ngwane, who moved onto the highveld south of the Vaal, and the Khumalo, or Ndebele as they were later known, who established themselves near the upper Vaal in what is now the south-eastern Transvaal. It is usually assumed that these groups had been driven out of the Natal-Zululand region by the expanding Zulu chiefdom, but recent research suggests that the realities were much more complex. They were squeezed out of a zone where conflict was intensifying as a result not only of the expansion of the Zulu but also of the Ndwandwe, and, possibly, of slave-raiding from Delagoa Bay (see below).

The intrusion of the Hlubi and the Ngwane certainly contributed to the widespread raiding and social disruption which took place on the southern highveld in the 1820s. But it is impossible to understand the origin of these upheavals without taking into account the extent to which the region north of the Orange had already been destabilized by the expansion northwards of the Cape colonial frontier zone. The Hlubi and Ngwane did not burst onto a peaceful and stable scene, as numbers of history books imply: they were pushed into an area where political tensions had been rising for many years.

This point is central to the scenario sketched out by Julian Cobbing in the article of 1988 which touched off the current debates on the mfecane. In various forms, the point had been hinted at a number of times before by other authors, but invariably in conjunction with assumptions about a Zulu- centred mfecane/difaqane which prevented it from being devel- oped further. Thus, twenty-five years ago, Martin Legassick

37. Legassick, 'The northern frontier', in Elphick & Giliomee, eds., Shaping, pp. 369, 375. 38. P.J. van der Merwe, Die Noordwaartse Beweaina van die Boere voor die Groot Trek (1770-1842). The Hague, 1937, p. 118. 14 could argue that the political instability of the 1820s in 'Transorangia' was a product not simply of 'the Difaqane' but also of 'the comparable, though usually less cataclysmic, tensions engendered by the spread of the southern frontier'.39 At the same time, William Lye made the point that 'The Sotho people faced a more lasting, and in some ways more destruc- tive, threat than the Difaqane in the plundering raids of the numerous communities of Korana and Griqua'.40 Wagner suggested that the raids of Coenraad de Buys probably contributed to the climate of 'dislocation and apprehension7 in which the 'Difaqane' was to 'thrive'.41 I have already cited Huffman's recognition that raiding from the south had made an important contribution to the instability of the southern Tswana region in the later eighteenth century.

It is only now, after Cobbing's intervention, that historians are beginning to recognize the full significance of this factor. Elizabeth Eldredge, who has sharply criticized Cobbing on other counts, accepts basic aspects of the case which he makes in this connection, and her own forthcoming essay on the slave trade across the northern frontier serves to reinforce it.42 While attempting to appropriate them to a revised concept of the mfecane, even John Omer-Cooper accepts the validity of parts of Cobbing's line of argument.43 Merinda van Zyl, who stands outside the narrow circle of mfecane debaters, has independently drawn on Legassick's and Cobbing's work to argue that the upheavals in the interior, which she continues to call the Difaqane, were the products not of 'Zulu refugees' but of raiding and counter-raiding set in motion by the expansion of the Cape colonial frontier.44

Before the impact of the thrust of the Hlubi and Ngwane into the southern highveld can be properly understood, the whole history of the region in the 1820s - and before, for that matter - needs to be rethought from scratch. A revised version, I would insist, should concentrate on trying to

39. Legassick, 'The Griqua', unpubl. thesis, p. 327; also p. 377. 40. W. Lye, 'The Sotho wars in the interior of South Africa, 1822-1837', unpubl. Ph.D. thesis, UCLA, p. 111. 41. Wagner, 'Coenraad de Buys', p. 5. 42. E. Eldredge, 'Sources of conflict in southern Africa, c.1800-30: the "mfecane" reconsidered'. Journal of African History, vol. 33 (1992), pp. 15-16, 25; Eldredge, 'Slave- raiding across the Cape frontier', in press. 43. J.D. Omer-Cooper, 'Has the Mfecane a future? A response to the Cobbing critique'. Journal of Southern African Studiesr vol. 19 (1993), pp. 286, 288, 290. 44. M.E. van Zyl, 'The quest for cattle, the firearm frontier and the Difaqane', Acta Academica. vol. 24 (4), (1992), pp. 58-73. 15 provide a comprehensive politics of the period, and not merely another repetitive account of fights and migrations. Current versions of this history are descended primarily from syntheses originally produced in the 1960s and 1970s, particularly those of William Lye,45 and for the most part have broken little new ground. An essential starting-point for a rethought history must be a critical re-examination of what has long been the standard source-book on the subject, Ellen- berger's History of the Basuto. first published in 1912. The revealing, if brief, comments on the production of this work made by Sanders need to be carried much further.46 A revised history would, among other things, need to assess the impact on the Sotho-speaking chiefdoms of the Caledon region of the increasing pressures exerted in the 1820s not only by the Griqua to the west and the Dutch to the south, but also by expanding powers like the Ndebele kingdom to the north and the Zulu kingdom to the east.

Also needed is a re-examination of the significance of the so- called battle of Dithakong in 1823. A great deal has been made by generations of colonial and colonial-influenced writers of this event, when a force of armed and mounted Griqua, mobilized at the behest of the missionary Robert Moffat, beat off/attacked a group of raiders from the highveld who had crossed the lower Vaal into the territory of the Tlhaping. A lot of ink has recently been spilt over Julian Cobbing's re-interpretation of the affair, in the course of which he provocatively (and not convincingly, in the eyes of numbers of readers, including myself) characterized Moffat as a slave-raider. Whether the episode was a major historical event, or whether it was an encounter which has received disproportionate publicity because of the presence of literate observers at the scene are questions which have not yet been answered.

In the Caledon valley, the evidence seems clear enough, the dominant power by the mid-1820s was the intrusive Ngwane chiefdom under Matiwane. In the face of raids made by the Griqua and other groups, the Ngwane moved away south-eastwards into the territories of the Thembu. As is well-known, in 1828 they were attacked and broken up by a force of British troops supported by a large army of Thembu, Gcaleka, Mpondo, and Cape settlers, all eager to seize cattle and people. This episode

45. Lye, 'The Sotho wars', unpubl. thesis, ch. 2; W. Lye, 'The Difaaane: the mfecane in the southern Sotho area, 1822-24', Journal of African History, vol. 8 (1967), pp. 107-31. See also Legassick, 'The Griqua', unpubl. thesis, ch. 7. 46. P. Sanders, Moshoeshoe: Chief of the Sotho, London, 1975, p. 332. I have not yet managed to get hold of E. Maloka, 'Missionary historiography and ethnography: Casalis, Ar- bousset and Ellenberger, and the history of nineteenth- century Lesotho', unpubl. Hons, thesis, UCT, 1988. 16 has become the subject of some controversy;47 if it is to be properly understood, it needs to be seen in the context of a thorough re-examination of British policy on the Cape eastern frontier at this period, as well as of the politics of the contemporary Xhosa-speaking chiefdoms. The migration and destruction of the Ngwane chiefdom opened the way for the emergence of two new rival powers on the Caledon. These were the Tlokwa chiefdom under Sekonyela in the north, and, further south, the Mokotedi chiefdom under Moshweshwe. Because of his later political successes in building up the Sotho kingdom, Moshweshwe has earned much more attention from historians than have other leaders of the period. Needed now is a history of the southern highveld in the 1820s and 1830s which avoids the common Sotho-centric teleologies, and provides a critical explanation of how Mokotedi domination of the Caledon region came to be established. A revised history should not lose sight of the fact that to the south-west, another focus of power, if a more loosely organized one, was emerging in the form of the Griqua- dominated polity of Adam Kok II. Kok had been encouraged by missionaries to settle at Philippolis near the Orange river in 1826. He exercised a weak authority over a turbulent follow- ing of Griqua, Kora, Bushmen, Tswana and Sotho. Further down the Orange, Andries Waterboer ruled another Griqua polity centred on Griquatown. Other small communities of Griqua and Kora lived along the middle Orange, the lower Vaal, and the Caledon.48 The roles played by all these groups need to be considered in an integrated regional history.

The history of the region north of the Vaal river in the 1820s and 1830s also needs to be comprehensively rethought. The established academic analyses of Lye, Morrow, Cobbing and Ras- mussen all belong to the period of the 1960s and 1970s, and little has been done since to subject their findings to critical analysis, or to re-interrogate the sources on which they are based. The rise to dominance of the Ndebele kingdom under Mzilikazi will no doubt remain the central theme of a revised history, but it needs to be placed in a much broader context than has hitherto been done. A future political history of the northern highveld in the period under review would need to try to establish a clearer picture than we have at present of the politics and economics of the bewildering succession of raids and counter-raids among the chiefdoms of the region. It would also need to provide an integrated analysis of the impact on the region's affairs of cattle raids by Griqua and

47. Cobbing, 'The mfecane as alibi', pp. 500-3, 507-9; J. Peires, 'Matiwane's road to Mbholompo: a reprieve for the Mfecane?' in Hamilton, ed., The Mfecane Aftermath, in press. 48. Marais, Cape Coloured People, pp. 36-42; Legassick, 'The Griqua', unpubl. thesis, ch. 8; Ross, Adam Kok's Griquasr ch. 3. 17 Kora from the south and by the Zulu from the east, of slaving by the Portuguese and their African allies from Delagoa Bay, of the expansion of the Pedi and Swazi chiefdoms to the east, and of the upheavals which were taking place north of the Limpopo river in the 1830s. The east: trade and politics. 1760-1820 Over the last fifteen years or so, the history of the territories east of the highveld in the period from the mid- eighteenth century to the 1830s has been subjected to more by way of critical revision than has that of the interior regions. The empirical work done by researchers like David Hedges, Jeff Guy, Philip Bonner, Peter Delius, Carolyn Hamilton and myself has largely supplanted earlier writings. The re-evaluations made by Carolyn Hamilton and myself of what has since 1929 been the standard work on the Natal region in this period, A.T. Bryant's Olden Times in Zululand and Na- tal,4^ have opened the way for thorough-going reconceptualiza- tions of its history. The publication since 1976 of important new source material in the volumes of the James Stuart Archive has provided a basis for fresh research work to be done well into the future.50

In the territories east of the highveld, patterns of political change cannot be traced even in outline before the mid- eighteenth century. Evidence from recorded oral traditions synthesized by the researchers mentioned above suggests that in the second half of the century a number of chiefdoms over a wide region were beginning to expand, and frequently to clash with one another and with their smaller neighbours. As already mentioned, Smith and Hedges have demonstrated that long-distance trade through Delagoa Bay was growing rapidly after about 1760, with ivory going out and cloth, beads and metal coming in. It is a likely hypothesis, in my view, that competition for this trade was a major factor in increasing the level of political conflict in the region. There is some consensus among researchers that, by the end of the eighteenth century at least five enlarged chiefdoms had emerged in the hinterland of the bay. To the north-west, near the Olifants river, Delius indicates, the Maroteng (or Pedi)

49. J. Wright & C. Hamilton, 'Traditions and transformations: the Phongolo-Mzimkhulu region in the late eighteenth and early nineteenth centuries', in A. Duminy and B. Guest, eds., Natal and Zululand from Earliest Times to 1910: a History. Pietermaritzburg, 1989,pp. 49-82; J. Wright, 'A.T Bryant and "the wars of Shaka"', History in Africa, no. 18 (1991), pp. 409-25. 50. C. Webb & J. Wright, eds., The James Stuart Archive of Re- corded Oral Evidence Relating to the History of the Zulu and Neighbouring Peoples. 4 vols. so far published, Pietermaritzburg, 1976-1986. 18 chiefdom was establishing control over the trade routes from Delagoa Bay to the interior.51 At the bay itself, Smith and Hedges make clear, the Tembe and Maputo chiefdoms were compe- ting for control over trade routes to the west and south.52 Further south, the Ndwandwe and Mthethwa chiefdoms were becoming rivals for domination of the region north and south of the White Mfolozi: Hedges suggests that struggles for control of trade to the bay from the south may have been a factor in their conflict.53

As these chiefdoras expanded, numbers of groups on their margins were pushed out. Among them were the Dlamini, who moved from the region south of Delagoa Bay into what is now Swaziland.54 They formed the nucleus of what later became the Swazi kingdom. Other groups, Hamilton and I have argued, began coming together for defensive purposes.55 Among them were enlarged chiefdoms like those of the Hlubi and Qwabe in what is now Natal.56 The expansion of the Qwabe, my own researches reveal, in turn pushed other chiefdoms southwards along the coast. Some migrant groups ended up in the northern Transkei, where their arrival was probably an important factor in stimulating the expansion of the Mpondo chiefdom.57

As larger chiefdoms started to emerge, whether in response to rivalries over trade or for defensive purposes, conflicts between them seem to have become more intense. Hedges argues

51. Delius, The Land Belongs to Us. pp. 13, 15-19. 52. Smith, 'Struggle for control of southern Mocambique', unpubl. thesis, pp. 223-4; Hedges, 'Trade and polities', unpubl. thesis, ch. 5. 53. On the expansion of the Ndwandwe, see Hedges, 'Trade and polities', unpubl. thesis, pp. 155-65. On the expansion of the Mthethwa, see Hedges, 'Trade and polities', pp. 177-86; also C. Hamilton, 'Ideology, oral traditions and the struggle for power in the early Zulu kingdom', unpubl. M.A. thesis. University of the Witwatersrand, 1986, pp. 102-19. 54. P. Bonner, Kings. Commoners and Concessionaires: the Evolution and Dissolution of the Nineteenth-Century Swazi State. Johannesburg, 1983, pp. 9-12. 55. Wright & Hamilton, 'Traditions and transformations', p. 65. 56. J. Wright, 'The dynamics of power and conflict in the Thukela-Mzimkhulu region in the late 18th and early 19th centuries', unpubl. thesis. University of the Witwatersrand, 1990, pp. 34-9. 57. Wright, 'Dynamics of power', unpubl. thesis, pp. 40-9; J. Wright, 'Political transformations in the Thukela-Mzimkhulu region of Natal in the late eighteenth and early nineteenth centuries', in Hamilton, ed., Mfecane Aftermath, in press. 19 that the decline of the ivory trade and the rise of a trade in cattle at Delagoa Bay in the 1790s further exacerbated political tensions.58 Over the years, numbers of historians have repetitively argued that a severe drought, which is commonly supposed to have taken place in the very early years of the nineteenth century, was another factor in the escalation of political struggles in the Natal region. The sources of this notion have never been critically examined, and it is likely, in my view, to turn out to be one of those enduring myths which lodge so easily in the fabric of mfecane-informed history. Which is not to say that climatological factors were not important in the history of the period: they undoubtedly were, but until in-depth research on the subject has been published, it is very difficult to make more than vague and gen- eralized statements on it.

Another factor in the politics of the Delagoa Bay hinterland by the 1810s may possibly have been the growth of a slave trade at the bay. This idea was first argued by Julian Cobbing, who, in his 1988 article sees this trade as having been significant since well back in the eighteenth century.59 His argument has been heavily criticized by Elizabeth Eldredge, who insists that the existing evidence indicates that the slave trade at Delagoa Bay would have been of no importance before about 1823.60 My own opinion is that Cobbing heavily overstates his case, and that the trade at Delagoa Bay is unlikely to have taken off before the rise of exports to Brazil after 1810. I would also argue that some of the evidence cited by Eldredge is more ambiguous about the timing of the origins of the trade than she allows for, and it may possibly have been expanding, if on a small scale, before 1820. In a more recent essay, Eldredge has in fact slightly modified her argument in this direction.61

One of the reasons why I want to hold on to the idea of a pre-1820 origin for the slave trade is so that it can be deployed as a possible explanation for the intensification of conflict which took place south of the Phongolo in, it seems, the late 1810s. Recorded traditions suggest that the Ndwandwe were acting with an unusual degree of aggression towards their neighbours at this time: as I have argued elsewhere, this may have been in part because they were coming increasing pressure from the effects of slaving activities to the north.62

58. Hedges, 'Trade and polities', unpubl. thesis, pp. 145-52. 59. Cobbing, 'The mfecane as alibi', pp. 503-7. 60. Eldredge, 'Sources of conflict', pp. 3-15. 61. E. Eldredge, 'The Delagoa Bay hinterland: politics, trade, slaves and slave-raiding before 1850', in Eldredge & Morton, eds., Slavery in South Africaf in press (my thanks to Elizabeth Eldredge for providing me with a draft of this chapter). 62. Wright, 'Dynamics of power', unpubl. thesis, pp. 211-12; Wright, 'Political transformations', in Hamilton, ed., Mfecane 20

As is well known, the outcome of the conflicts was the estab- lishment of Ndwandwe domination over the region from the Phongolo south towards the White Mfolozi. The common view is that the Ndwandwe were soon afterwards defeated, and their chiefdom broken up, in an attack by the Zulu under Shaka, who had been consolidating their power in the territory south of the Mfolozi. This view stands in need of re-examination. Hamilton's and my research into the early history of the Zulu kingdom indicate that it is unlikely that at this stage it had anything like the power single-handedly to attack and destroy the better-established and much more powerful Ndwandwe polity. There is little doubt that the Ndwandwe chiefdom broke up soon after a Ndwandwe raid southwards had been beaten off by the Zulu, but the view that the Zulu then proceeded to attack and destroy the Ndwandwe chiefdom is based very much on the traditions of the Zulu rather than those of the Ndwandwe. Recently recorded Ndwandwe traditions, which themselves need critical assessment, point by contrast towards internal political tensions as the prime cause of the chiefdom's break-up.63

Subsequently the Ndwandwe ruling house and its adherents under Zwide moved into southern Swaziland and then into the eastern Transvaal, where, it seems, they attacked and broke up the Pedi chiefdom.64 The Khumalo under Mzilikazi moved to the south-eastern Transvaal, the Msane to eastern Swaziland, and the Jele (or Ngoni) and Gaza to the Delagoa Bay region.65 Another largely unexamined assumption of mfecane-informed history is that all these groups fled as 'refugees' from the Zulu under Shaka. This view neglects the possible impact of other factors, like the expansion of the Delagoa Bay slave trade, and does not allow for the possibility that these groups may have moved on their own initiatives.

The break-up of the Ndwandwe polity and the migration of these groups opened the way for the Zulu to extend their rule north of the White Mfolozi. Contrary to the common view, it is unlikely that at this stage their domination extended beyond the Mkhuze river, for their hold on the polity which they had

Aftermath, in press. 63. For exmple, see the archives of the Swaziland Oral History Project (University of the Witwatersrand), Ndwandwe history, box 1, evidence of John and Bongani Nxumalo, interviewed by Dumisa Dlamini for Swaziland Broadcasting Ser- vice, 3 June 1983, notebook 1, pp. 7-9; Ndwandwe history, box 1, evidence of Bongani Nxumalo interviewed by the same, 10 June 1983, notebook 3, pp 42-58; Ndwandwe history, box 1, evidence of Mafutha Mazibuko interviewed by Philip Bon- ner at Nhlunguyavuka, 11 June 1970, typescript, p. 1. My thanks to Ronette Engela and Carolyn Hamilton for guiding me towards these sources. 64. Delius, The Land Belongs to Us. pp. 19-24. 65. Wright, 'Dynamics of power', pp. 189-91. 21 established was still tenuous, and their concerns on this frontier were essentially defensive. To the north was a zone of instability where the intrusive Msane, Ngoni and Gaza rivalled one another, as well as the Portuguese, for domination. To the north-west the Ndwandwe were building a new polity and re-establishing some of their former power.^6 In this volatile situation, there was nothing to indicate which of these groups, including the Zulu, would emerge as dominant.

The east: the consolidation of new states in the 1820s and 1830s By the mid-1820s the political scene east of the highveld was very different from what it had been in the 1810s. The two major political units in the region were the emerging states or kingdoms ruled by the Ndwandwe under Zwide in the eastern Transvaal and by the Zulu under Shaka south of the Mkhuze river. In the literature, the size and power of the Zulu kingdom at this time has been heavily exaggerated and mythologized. To a large extent this is because it survived the upheavals of the 1820s to become the dominant power in the region, and to take an important place in history books and in oral traditions. But, as Russell Martin and Carolyn Hamilton have argued in different ways, much of its reputation as fero- cious military power was constructed by white and black writers and story-tellers in later years, particularly in the decades after its defeat by the British in the Anglo-Zulu war of 1879.67 In rethinking the history of the early Zulu kingdom, we need to be very much on guard against uncritically accepting received opinion, academic and non-academic alike.

Research done by Carolyn Hamilton and me since the early 1980s, in which we build particularly on the earlier work of David Hedges, points towards conclusions that in many respects modify or overturn those in conventional accounts.68 In the first half of the 1820s the main strategic aim of the Zulu leaders was to safeguard their northern borders against possible raids by groups from the turbulent regions beyond. To this end, Shaka established a number of strong settlements among the chiefdoms south of the Mkhuze river. Contrary to the common view, the new Zulu state did not have the military

66. Wright, 'Dynamics of power', unpubl. thesis, pp. 187-93. .67. R. Martin, 'British images of the Zulu, c.1820-1879', unpubl. Ph.D. thesis, University of Cambridge, 1982; C.A. Hamilton, 'Authoring Shka: models, metaphors and historiography', unpubl. Ph.D. thesis, Johns Hopkins University, 1993. 68. C.A. Hamilton, 'Ideology, oral traditions and the struggle for power in the early Zulu kingdom', unpubl. M.A. thesis. University of the Witwatersrand, 1986; Wright, 'Dynamics of power', unpubl. thesis. 22 or political means to extend its direct rule further north.69 Nor was it strong enough to pursue a policy of active expansion to the south. In this direction, Shaka's main concern was to establish domination over the chiefdoms on both banks of the Thukela river and in the coastlands to the south. Though Zulu forces made several raids into the teritories south of the river, they did not, as is commony believed, 'devastate' the region. The political upheavals which it suffered in the 1820s were caused primarily by the depredations of groups migrating away from the zone of conflict north of the Thukela.70

Though the Zulu leaders sought to establish a system of strong central government, their kingdom remained only loosely united. Like other African kingdoms, it was an amalgamation of separate chiefdoms under a common overlord. Throughout Shaka's reign, the Zulu faced strong opposition from some of the chiefdoms subject to them. In addition, Shaka had to deal with the hostility of rival factions within the Zulu ruling house.71 In the last years of his reign, he turned more and more for support to the small parties of British traders which, from 1824, he had allowed to operate from Port Natal. He looked to them to supply him with prestige goods from the outside world, to open up commercial and diplomatic relations with the Cape, and to assist him against his enemies inside and outside the kingdom.72

In contrast to the case of the successful Zulu polity, we know very little of the history of the Ndwandwe. The defeat and final break-up of their kingdom in 1826 (see below) meant that most of its history became lost to sight, and in most accounts of the 1820s the Ndwandwe remain little more than a shadowy presence in the eastern Transvaal. But the surviving scraps of evidence suggest that by the early 1820s their chiefdom was emerging as the dominant power in the region. Delius argues that it was strong enough to defeat the established Pedi polity,73 and evidence adduced by Bonner indicates that the Ndwandwe attacked the expanding Dlamini chiefdom in what is now Swaziland on several occasions. While some of these

69. Wright, 'Dynamics of power', unpubl. thesis, pp. 188-93. 70. Ibid., chs. 5, 6; Wright, 'Political transformations', in Hamilton, ed., Mfecane Aftermath, in press. 71. On internal opposition to Shaka's rule, see Hamilton, 'Ideology', unpubl. thesis, pp. 184-6, 362-3; Wright, 'Dynamics of power', unpubl. thesis, pp. 253, 281-3, 353-5, 367-9. 72. Wright, 'Dynamics of power', unpubl. thesis, ch. 7; C. Hamilton, '"The character and objects of Chaka": a reconsideration of the making of Shaka as "mfecane" motor', Jnl. Afr. Hist.. vol. 33 (1992), pp. 41-57. 73. Delius, The Land Belongs to Us. pp. 19-24. 23 attacks were made when the Ndwandwe were still living in the vicinity of the Phongolo river, others may have taken place after they had moved to the eastern Transvaal.74 Also in this later phase, they may have attacked the Khumalo or Ndebele under Mzilikazi in the upper Vaal region.75 If so, pressure from Ndwandwe expansion may have been a factor in Mzilikazi's decision to shift further westward into the southern Transvaal (see above). Another factor, as Cobbing suggests, may have been the growth of slaving in the Delagoa Bay hinterland.76

By 1826 the expanding Ndwandwe kingdom seems to have been pressing hard on the north-western borders of the Zulu kingdom. If the evidence of the British trader Nathaniel Isaacs is anything to go by, Shaka viewed the growth of Ndwandwe power with considerable alarm.77 With the help of a small party of gunmen from the British trading posts at Port Natal, the Zulu king launched an attack on the Ndwandwe, defeated them, and captured large numbers of cattle. This episode is usually seen in the literature through the eyes of Henry Fynn, a British trader who participated in the attack and has left a brief account of it. He portrays the Zulu attack as a response to an invasion on the part of the Ndwandwe: my own reading of the evidence suggests that Shaka took advantage of the presence of the British with their firearms to make an attempt to destroy his main rival power.78 In the event, his initiative succeeded. After its defeat, the Ndwandwe kingdom fell apart. Some groups submitted to Shaka, while others went off to give their allegiance to the leaders of other emerging states in the region.

The break-up of the Ndwandwe left the Zulu kingdom as the strongest state east of the Drakensberg. Two years later, in 1828, Shaka was assassinated in a 'palace coup' led by his brother Dingane, who replaced him as ruler. The motives behind this event, and its wider significance, have recently been the subject of several new and differing interpretations.7^ The emergence of these differences points to the

74. Bonner, Kings. Commoners and Concessionaires, pp. 27-9; and Bonner, pers. comm. 75. This is my interpretation of references to Ndwandwe- Ndebele hostilities in T. Arbousset, Narrative of an Ex- ploratory Tour to the North-East of the Colony of the Cape of Good Hope. Cape Town, 1968 (1st publ. Cape Town, 1846), pp. 185, 294. 76. Cobbing, 'The mfecane as alibi', p. 506.

77. N. Isaacs, Travels and Adventures in Eastern Africaf Cape Town, date (ist publ. London, 1836), p. 71. 78. See the analysis of this event in Wright, 'Dynamics of power', unpubl. thesis, pp. 338-44. 79. Cobbing, 'The mfecane as alibi', pp. 513-19; Wright, 'Dynamics of power', unpubl. thesis, pp. 367-71; Hamilton, '"Character and objects of Chaka1", JAH, vol. 33 (1992), pp. 57. 24 pressing need for fresh research, drawing on the relatively plentiful evidence in the James Stuart Archive and in Cape archival documents and newspapers, into the relationship between the internal politics and foreign policies of the Zulu kingdom in the latter part of Shaka's reign and the early years of Dingane's. On this latter period, all that can be said here is that, like his predecessor, Dingane faced considerable resistance from powerful chiefs inside the kingdom, but, by eliminating a number of Shaka's closest supporters, and by making frequent presents of cattle to important political and military leaders, he was for a time able to keep opposition in check.80 Like Shaka, he was anxious to expand trading relations with the British at the Cape. He also encouraged the development of commerce with the Portuguese at Delagoa Bay, and in 1833 went as far as making an attack on them in order to keep open the trade route to the bay.81

Another effect of the collapse of the Ndwandwe kingdom in 1826 was to leave the way open for the expansion of the Dlamini or Swazi chiefdom under Sobhuza, and the re-emergence of the Pedi chiefdom under Sekwati. In the 1830s, Bonner tells us, Sobhuza was extending his rule over much of Swaziland and adjoining regions of the eastern Transvaal. By 1836 his chiefdom was firmly enough established to survive an attack by a large Zulu force sent against it by Dingane.82 To the north-west, in the region of the Olifants river, Delius's research shows, Sekwati was able to rebuild Pedi ascendancy through a combination of clever diplomacy and successful raiding of cattle from smaller groups. He was also able to attract numbers of refugees from the upheavals caused in neighbouring territories by the raids of the Swazi, the Portuguese, and the 8^

The history of the eastern Transvaal-Swaziland-Zululand region in the 1820s and 1830s will not be better understood until it has been more comprehensively integrated with that of southern Mozambique. To achieve this, anglophone researchers in the region will need to take more account than they do at present of the researches of their colleagues in Mozambique. One important theme which needs to be researched and written from a more integrated perspective is the expansion and impact of the Delagoa Bay slave trade. Whatever the arguments about when the trade originated, commentators seem to be agreed that

80. On the early years of Dingane's reign see Hedges, 'Trade and polities', unpubl. thesis, pp. 242-4; Wright, 'Dynamics of power', unpubl. thesis, pp. 372-5. 81. G. Liesegang, 'Dingane's attack on Lourenco Marques in 1833', Jnl. Afr. Hist.. vol. 10 (1969), pp. 565-79; Hedges, 'Trade and polities', unpubl. thesis, pp. 242-4. 82. Bonner, Kings. Commoners and Concessionaires, pp. 39-41.

83. Delius, The Land Belongs to Usf pp. 26-30. 25 by the mid-1820s it was growing in importance.84 But because its significance has only recently come to be appreciated by historians, as yet very little is known either about its extent or about its effects on the societies of the region. Another major theme has to do with the processes by which the Gaza kingdom under Soshangane emerged by the late 1820s as the dominant power in southern Mozambique. For scholars brought up on now dubious ideas about the extent and power of the 'Zulu empire', it would be of particular interest to learn how it came about that the sway exercised by Soshangane eventually extended over much of Mozambique as far north as the Zambezi river, a much larger region than that dominated by any other chief in southern Africa.

Conclusion Readers will have noticed that, in outlining a new approach to explaining the processes of political change which took place in the interior and eastern regions in the period from the 1760s to the 1830s, I nowhere need to make use of the concept of the mfecane. This is not because I am looking for a new name for it; it falls away because I adopt an analytical focus which is quite different from that of mfecane-informed history. The latter almost invariably assumes, on the basis of minimal evidence, that the most significant political changes of the era were concentrated into the 1820s and 1830s, and that they were set in train primarily by the expansionism of the Zulu kingdom. These notions are incompatible with the argument which I put forward here.

As I see it, two separate and expanding zones of political instability were emerging from about the 1760s onwards, one on the south-east fringes of the Kalahari and the other in the hinterland of Delagoa Bay. They were the product of the meshing of European colonial and commercial expansion with indigenous political conflicts. By the 1820s these zones were beginning to overlap; the result was an intensification of conflict, particularly in the interior regions, and the emergence of a number of relatively large and centralized polities. These states, as we may call them, were still consolidating themselves when, in the later 1830s, the migration of parties of Boers and their retainers from the Cape into the territories across the Orange and Vaal inaugurated a new era of violence and instability in much of southern Africa.

84. Eldredge, in Hamilton, ed., Mfecane Aftermath, in press.